ICUMSA News n°18 – 1993

Message from the President

In the January issue of International Sugar Journal, Dr James C.P. Chen commented on NIR polarimetry [1], generally endorsing its adoption along the recently announced lines that Domino Sugar (formerly Amstar) propose, with respect to suppliers of raw sugar to that company. As President of ICUMSA and the former Referee for “Polarisation of Raw Sugars”, I would comment that I think that this move by Domino and the endorsement by Dr Chen are premature and usurp the role of our Commission to make recommendations in such matters for the benefit of all parties in industry and commerce.

In the May 1992 issue of ICUMSA News, the Commission heard from its Referee for “Polarimetry and Quartz Plates” on this subject. Dr Emmerich stressed that Bünnagel’s formulae “should not be used where measurements of higher accuracy are required”. He cited the order of magnitude of error in using those formulae as 0.06 °Z, measured in this author’s laboratory [2, 3].

When one recalls that it was the United States National Committee for sugar Analysis (USNC) that objected to the adoption of a 0.03 % change in the 100° point of the Sugar Scale in 1974, it seems unduly hasty for Dr Chen to advocate a change in the method before the polarimeter it employs is properly standardized. That Domino Sugar claims this haste derives from a sense of urgency with respect to obviating the use of lead compounds when, on the other hand, they have done nothing to adopt the 1986 Recommendations concerning the ICUMSA methods seems inconsistent.

It is claimed by Altenburg and Chou that on average there is little difference in results between the old and new method. This may be true, but it does not take into account that the old method is the pre-1986 method using the 0S Sugar Scale and deducting 0.1 ° from the polarisation result and that standardization of the NIR polarimeter is possibly 0.06 °Z in error. Dr Emmerich calculates that the +0.005 difference is really -0.066 °Z with the NIR method giving the lower result.

If one looks at Altenburg and Chou’s Table showing the difference in result according to cargoes [4], we observe that some 5 % are reported with a difference greater than 0.1 degrees. Even without calculating what it means on a common scale and with the NIR polarimeter properly calibrated, it shows that the introduction of the NIR polarimetric method produces some winners and losers and that it is not valid to say that since on average the new result is about the same as the old, we have a valid basis for substituting the new for the old.

As President, I would say to the USNC and its Chairman, this Commission should have a role in recommending the adoption of a new polarimetric sucrose determination method because of its ramifications for industry and commerce. I would encourage you to use your influence in persuading Domino Sugar that it is being unnecessarily hasty in promoting this new NIR polarimetric method at this time.

Physical methods for analysing sugar products

by K.J. Rosenbruch, Chairman of the German National Committee,

Referee of Subject 13

Physical methods of measurement for the analysis of sugar products have been used for a long time and, as is well known, the standardization of polarimetric methods was one of the first tasks of ICUMSA founded nearly 100 years ago.

Chemical analysis of the various substances, i.e. the different kinds of sugars on the accompanying substances in the sugar products is, of course, indispensable. However, physical methods, if applicable, usually allow a much higher accuracy to be achieved in measuring the concentration than do chemical methods of analysis.

The physical methods most frequently applied in sugar analysis are:

Refractometry. polarimetry and densitometry. The quantities measured are the refractive index, the angle of rotation and the mass per volume.

These are not parameters specific to sugars; they are determined using aqueous solutions, and when modern measuring instruments are used, the refractive index can be measured with an accuracy of ± 1 ´ 10-6 and the angle of rotation with an accuracy of ± 1 ´ 10-4 degrees. The accuracy achievable in density measurements is ± 1 ´ 10-6 g/cm3 or ± 1 ´ 10-3 kg/m3. The high accuracy inherent in these physical methods can be achieved only if some parameters are exactly known and constant. These are the temperature, and, for optical measurements, the temperature and the wavelength of the light employed. Physical measurements require that measurements be made on two solutions, one being an aqueous solution of a product as it is obtained during the manufacture of sugar; the other is an aqueous solution of purest sucrose or of one of the other purest sugars which are of interest, such as glucose, fructose, invert sugar etc. Only if the content of components other than the sugar in question is known can the concentration of sugar in the product under investigation be deduced from the physical measurements of the refractive index, the optical rotation or the density.

In addition to sucrose, technical solutions contain other sugars and non-sugars, which influence the refractive index, the density and the optical rotation in different ways and result in the deviations from the values of pure solutions. A comparison of the results obtained for the apparent sucrose content of a solution using all three methods can provide information about the character and concentration of the other components in the solution.

ICUMSA has officially adopted tables specifying the refractive indices and densities of these highly purified aqueous sucrose solutions. The optical rotation has been accepted by the new definition of the International Sugar Scale. The physical quantities have also been given for temperatures differing from the standard temperature of 20 °C and for wavelengths of light differing from the wavelength of 589.3 nm of the yellow Na lines (mostly used for refractometric measurements) and of 546.1 nm of the green mercury line (used in polarimetry measurements) [1-4]. The accuracy of these fundamental values and reference tables is considerably greater than that required in practical applications. Instruments for measuring the concentration of aqueous solutions of sugars are provided with scales based on these values. In polarimeters, the composition in % at a given concentration is measured instead of the angle of the optical rotation. Refractometers measure the concentration of pure aqueous sucrose solutions instead of the refractive index and in densitometers, as well, the concentration of pure aqueous solutions is measured instead of the density.

These instruments can be calibrated and checked by means of quartz plates (polarimeter), test glasses and distilled water (refractometer, densitometer).

It is not the task of ICUMSA to develop and improve these instruments, but for ten years now automatic instruments of high resolution and good reproducibility have been on the market for the three physical methods of analysis and these are gradually replacing the older manual instruments, whose operation requires visual manipulation by skilled and experienced personnel. Such automatic instruments were successfully used in several collaborative tests [5]. Details of construction, maintenance and instructions for use will be elaborated by ISO and OIML to guarantee the necessary accuracy and reproducibility of the results obtained with these instruments.

The main task of ICUMSA is to set the tables for (or define mathematical relations between) the concentration of aqueous solutions of highly purified sugars and the physical quantities, i.e. refractive index, density, polarisation. As to sucrose, the work being done covers refractive index, density and polarisation. For the monosaccharides mostly used, i.e. glucose, fructose and invert sugar, and mixtures of these, relevant tables for the refractive index have been Officially adopted. Work regarding the density tables for these monosaccharides is proceeding well in the PTB, in cooperation with the “Institut für landwirtschaftliche Technologie und Zuckerindustrie”.

In recent years the measurement of the optical rotation of aqueous solutions of raw sugars using near infrared radiation has become increasingly important. To avoid lead acetate clarification of dark solutions which is necessary to make measurements in the visible spectral range possible using a conventional polarimeter, near infrared radiation of 850 nm and above (e.g. 880 nm) is employed. This is because transmission of such light radiation is sufficiently high to allow the optical rotation to be measured without clarification. To proceed, the optical rotation of purest sucrose should be known at such wavelengths and for instrument calibration the optical rotation of quartz plates should be known as well. To-date, the optical rotation of sucrose solutions and of quartz is only exactly known and officially adopted by ICUMSA up to the wavelength of 644 nm (red Cd line). Extrapolation from 644 nm up to 850 nm and above without precise measurements are clearly affected by large uncertainties. The relevant measurements should therefore be done in well-equipped laboratories such as PTB, NIST or NPL. The German National Committee and the Referee of Subject 4 “Polarimetry and Quartz Plates”, Dr A. Emmerich, advocate that this task be done in cooperation between the PTB and the “Institut für landwirtschaftliche Technologie und Zuckerindustrie” in Braunschweig.

Apart from the above-mentioned measurements of the density of aqueous solutions of monosaccharides, which will also be carried out by the “Institut für landwirtschaftliche Technologie und Zuckerindustrie”, Braunschweig and the PTB, in my opinion, no further work on reference Tables will be necessary in the near future. However, all those interested in this work and willing to cooperate are invited to make proposals for an extension of these fundamental values and reference Tables to cover any other solutions and substances.


1 Proc. 17th Session ICUMSA 1978, 166

2 Proc. 20th Session ICUMSA 1990, 284

3 Proc. 19th Session ICUMSA 1986, 66

4 Proc. 20th Session ICUMSA 1990, 265

5 Proc. 20th Session ICUMSA 1990, 285

Clarification with aluminium sulphate for beet analysis

by Werner Mauch, Berlin Sugar Institute, Technische Universität,

General Referee for Subject 6: Beet

Since the 20th Session of ICUMSA, we have completed work on Recommendation 4 – Collaborative studies on the application of aluminium sulphate as a clarifying agent in the polarimetric determination of sucrose in sugar beet brei – with the committed support of Dr Steinle

of Südzucker. Eleven laboratories participated in the collaborative studies. These were located in Austria, Denmark, Germany, Italy and the Netherlands. The tests were carried out on five homogenised samples of frozen sugar beet brei. All the necessary precautions were taken before and during the dispatch of these samples in order to ensure that they did not thaw out during transport.

For five samples tested in the eleven laboratories, the range of mean sucrose contents, repeatability and reproducibility given below were computed (minimum and maximum values):

Mean values in °Z                                         16.843 ➔ 18.727

Repeatability value r (2.8 ´ sr)                      0.14 ➔ 0.30

Reproducibility value R (2.8 ´ sR)                0.22 ➔ 0.50

The Referee’s Report, which presented the result of the collaborative tests, was sent to the Chairmen of the National Committees by the President of ICUMSA, together with the comments of Mrs Mary An Godshall, Referee of Subject 3. By means of a letter ballot, the following recommendation was adopted:

“With reference to recommendation 4 of the 20th Session and after having evaluated the results of the collaborative tests, the determination of polarisation (°Z) in sugar beet brei samples using aluminium sulphate clarification according to the attached description should become an Official ICUMSA-method”.

A description of the method is attached to the Report for General Subject 6 of the 20th Session of ICUMSA of 1990 in the form of an Appendix. A further publication of the method that offers a most precise as well as detailed representation, is being prepared at present. This adoption does not mean that the former method, that has been employed for quite a long period of time – in which aqueous basic lead acetate was used as a clarifying agent – loses its Official ICUMSA Status at this time, for the reasons given below.

Until now, sufficient data regarding the influence of the aluminium sulphate clarification on the determination of potassium, sodium and α-amino nitrogen have not been available. Therefore, all parties that are interested in the creation of Official ICUMSA methods for the determination of these beet components, are called upon to conduct comparative tests with the beet material in their possession, employing two clarifying agents that are used for sucrose determination (aqueous basic lead acetate and aluminium sulphate). Since collaborative tests with the perishable brei samples involve great expenditure of time and costs, further results from comparative tests of individual laboratories should be awaited before we embark upon the collaborative tests once again.

In order to determine potassium and sodium, the flame photometric method is to be used after clarification with basic lead acetate as well as aluminium sulphate for comparison. Comparative tests are also required for the determination of α-amino nitrogen. These should be made by the Blue Number method, when using a solution of basic lead acetate as a clarifying agent (old method) and with the Ninhydrin and Fluorimetric methods when aluminium sulphate is employed (new method).

For such tests and those that have resulted from the last Recommendations adopted by ICUMSA, an appeal is made not only to the Associate Referees but also to anyone interested in further development of sugar beet analyses.

Inter-Agency Meeting, 6th November 1992, Budapest

Codex Alimentarius Committee on Methods of Analysis and Sampling

7th-l3th November, Budapest

by Brian Whitehouse (Cerestar Quality Assurance, Vilvoorde, Brussels, Belgium) and Roger Wood (Food Science Laboratory, Ministry of Agriculture, Fisheries and Food)

We represented, or looked after, ICUMSA’s interests at the above meetings.

1. Inter-Agency Meeting

The meeting was held in conjunction with the Codex Committee on method of Analysis and Sampling (CCMAS). It was attended by representatives from AOAC, ISO, IDF, NMKL, OIV, ICC, CEN, EOCQ and IFG (*) as well as from ICUMSA.

• AOAC Association of Official Analytical Chemists; ISO International Standard Organization; IDF International Diary Federation; NKML Nordic Analytical Committee; OJY Office International de la Vigne et du Yin; ICC International Association of Cereal Scientists and Technologists; CEN Comite Europeen de Normalisation; IFG International Federation of Glucose Industries

Information was exchanged on the inter-laboratory collaborative trials and the format of such trials that each of the above-mentioned International Organisations used or have underway at the present time. Most of the International Organisations working in the food analysis field have accepted the use of the 1 988 Harmonised Protocol on the Conduct and Evaluation of Collaborative Studies. It was reported that a further ISO/AOAC/IUPAC Harmonised Protocol on the Proficiency Testing of Analytical Laboratories has been developed and will shortly be published. In addition, it was noted that a third protocol on Quality Control is under development by ISO/AOAC/IUPAC.

There was significant discussion on the ownership of rights and copyrights for methods which are published by International Organisations. There were markedly different attitudes from different organisations, reflecting whether such organisations finance themselves by sales of methods or whether the preparation of such methods is a very minor part of the work of the organisation. In the latter case there is frequently no objection to methods being used by other organisations, provided due reference is made, as the potential loss of revenue is offset by the advantages that such organisations receive by the dissemination of their methods.

The problem has arisen because one organisation, CEN, frequently takes published methods of other international organisations and incorporates them into their own equivalent texts. For ICUMSA this should not be a problem; it is happening with those organisations which cover a wide range of commodities and not with those that are closely focused.

It was noted that testing of analytical kits is underway by an organisation set up by AOAC. This will validate claims made and, if satisfactory, will issue an endorsement valid for a year. This is in its early stage.


The 18th meeting of CCMAS followed on immediately after the Inter-Agency meeting. Various points of issue were discussed during that meeting, those which are of interest to ICUMSA members are given below.

Terms of reference

Amendments are proposed to make CCMAS a coordinating body for Codex with other international groups working on methods of analysis sampling and quality assurance systems for laboratories. Thus, CCMAS will be able to define procedures, protocols, guidelines and related texts for assessment of laboratory proficiency, as well as the methods of analysis and sampling aspects which it has concentrated on up to now.


It is proposed that a single Codex document on this subject will be prepared and that all sampling provisions be removed from individual standards. The first draft of a possible Codex Sampling Codex was outlined by the UK delegation.

Acceptance criteria for methods

There was much discussion on the information that must be supplied by Codex Commodity Committees when their methods are to be endorsed by the CCMAS. CCMAS is finding difficulty in endorsing methods in which insufficient information, mainly on the collaborative trials that have been carried out, is submitted. In particular, it runs into difficulty if collaborative tests have not been carried out because the method under consideration is “classical”. There will have to be agreement as to whether such methods may be “grandfathered”‘ into the Codex system. For ICUMSA methods which are undergoing collaborative trial at the present time, the CCMAS was very appreciative of the work which is being put into their validation. It is hoped that by the next Session of ICUMSA it will be clear as to what attitude Codex will take in future to methods which are “classical, traditional and which are used regularly in international trade” but “which have not been collaboratively tested”‘.

Criteria for evaluating acceptable methods

The UK delegation had prepared a paper on prescribing the use of method criteria rather than specific methods; it was attempting to state that the approach whereby any method may be used in a given situation provided that method meets specific criteria shall be considered. This is the EC approach in some areas and is thus important. The approach will not mean that existing methods are replaced but it will help to keep methods up to date, rather than their being enshrined on “tablets of stone”. It was unfortunate that the example used to illustrate this principle was about aflatoxins, unfortunate in that many people thought the paper was about aflatoxin methodology specifically, rather than as a means of evaluating acceptable methods for any analyte and thus to offer the analyst a wide choice of methods.

Methods for metallic contaminants

A series of general methods for heavy metals was listed and a document on sample preparation was also prepared. Ashing by various methods and the problems with these procedures were fully documented.

Methods which are required by Codex

In various Codex Standards there is still a need for methods to be incorporated. These methods were identified. In the case of sugars, the authors of this report were charged to try to clarify the situation in time for the next meeting. We will prepare a summary paper and present it at the next meeting. That paper will take the form of a general review of all the methods of analysis for sugars presently in the Codex Sugar Standards and identify where such methods need to be further produced.

Proficiency testing

A paper on the use of proficiency testing and a procedure for its conduct and statistical evaluation has been prepared by the UK. Most delegates were in agreement with the procedures being described, but France was a notable exception to this. It was agreed that the subject would be reconsidered at the next meeting. It should be appreciated by ICUMSA members that the use of proficiency testing will become a particularly important issue, especially now that at least some food enforcement authorities will prescribe its use for both legislative and accreditation purposes.


The meeting was useful as a number of new ideas were being floated in this international forum for the first time. This particularly applies to proficiency testing and to the use of method criteria. It is heartening to know that ICUMSA is well in advance of much of the work that it is doing, especially with respect to method validation.

Editor: Dr. R. Pieck, Donystraat 85, B-3300 Tienen, Belgium Tel.: +32 16 823 096 – Fax: +32 16 820 826 – Telex 27105